The Message of the Khirbat At-Tannør Reliefs

Home , Tutelary deity, Tyche, Uni (mythology)

Originalveröffentlichung in: Studies in the History and Archaeology of Jordan 10, 2009, S. 577-584

Robert Wenning Robert Wenning Institut für Altorientalische Philologie und Vorderasiatische Altertumskunde University of Muenster [email protected] The Message of the Khirbat at-Tannør Reliefs

The Nabataean sanctuary of Khirbat at-Tannør at Taurus, the Gemini, Cancer, Libra and a fragment the northern limit of Edom high above the Wådπ of Sagittarius and some Nike figures. The identifi- al-Óaså was excavated by Nelson Glueck in 1937 cation of the busts is facilitated by their crowning and became more famous by his book “Deities and attributes, such as the horns on the bust of Taurus. Dolphins”, published in 1965. People were fasci- nated by the richness and the style of the sculptural The Zodiac decoration of the sanctuary. Glueck’s book had a Basing on such attributes Judith McKenzie elabo- great impact on the research on the Nabataeans. A rated on the zodiac approach and gave a new in- few years ago Judith McKenzie, Sheila Gibson and terpretation of the so-called Dolphin Goddess from Andres T. Reyes published several important ar- Khirbat at-Tannør (Glueck 1965: 315-319 pls. 1-2; ticles on Khirbat at-Tannør (McKenzie, Reyes and McKenzie 2003: fig. 192). Nelson Glueck under- Schmidt-Colinet 1998; McKenzie 2001; McKen- stood this bust as a representation of the Syrian zie, Gibson and Reyes 2002; McKenzie, Reyes and goddess Atargatis. This relief was the corner-stone Gibson 2002; McKenzie 2003). Their re-examina- of his interpretation of the entire decorative pro- tion of the evidence has led to a new reconstruction gramme of Khirbat at-Tannør. McKenzie correctly of the sanctuary and a revised chronology, as well as demonstrated that the marine creatures in the hair new interpretations of the sculpture. I largely agree of the bust are clearly fish, and no dolphins, iden- with them and take their results as the starting point tical to the fishes in the zodiac (McKenzie 2001: for further considerations. My own study of astral 109; McKenzie, Gibson and Reyes 2002: 76-77). belief among the Nabataeans likewise focuses on The counterpart of the bust with the fish is a fe- the Khirbat at-Tannør sculptures and has resulted male bust with ears of grain in the hair (Glueck in some new ideas presented first in a lecture dur- 1965: 315-318 pls. 25-26, so-called Grain God- ing a conference of the Deutscher Palaestina-Verein dess; McKenzie 2003: fig. 191). McKenzie identi- 2006 and at the Washington ICHAJ 10. fied the two busts as zodiac signs, Virgo and Pisces, Khirbat at-Tannør and the nearby Khirbat adh- already suggested by Starcky 1968: 132. The busts Dharπ˙ were founded by the Nabataeans, but the decorated the outer pilasters of platform III (McK- monumental extension of the sanctuaries at both enzie 2003: fig. 188), which is about a century later sites took place in the first half of the second cen- than the main phase of the sanctuary. There are five tury AD (McKenzie, Gibson and Reyes 2002: other panels with busts above the zodiac busts, but 50, 72-73; Villeneuve and Muheisen 2000: 1535; they follow a more simple formula (Glueck 1965: cf. the lupa Romana Villeneuve 2002), that is the pls. 27-28). Clearly, the two zodiac panels are em- early phase of the new Provincia Arabia. The sub- phasized. This seems to be a citation of a composi- ject of the frieze of the Qaßr adh-Dharπ˙, excavated tion found among the sculptures of the older phase by François Villeneuve and Zeidun al-Muheisen, of the sanctuary. was recognized in 1999 by the bust of Libra with The famous zodiac (Glueck 1952, 1965: 395ff. a balance as busts of the zodiac figures in alterna- pls. 46-48; McKenzie 2001: 108-109; Rosenthal- tion with standing Nike figures (Villeneuve and Heginbottom 2001; McKenzie 2003: 186-191) is Muheisen 2000, 2003). Preserved are the busts of said to be contemporary with the older main phase

-577- Robert Wenning of the sanctuary or a little bit later for stylistic rea- beginning with Tishri on the right (Glueck 1965: sons, but whose original place in the sanctuary is 413-415). This presumed blending and integration unknown. The zodiac roundel with a bust of a fe- of two calendars and two different New Year fes- male deity in its middle is carried by a Nike (FIG. tivals side by side is questioned only by Starcky 1). In comparison with other zodiacs (cf. Gundel 1968: 232 (cf. Gundel 1992: 106, 222). McKenzie 1992) the arrangement of the signs at the at-Tannør 2001: 108-109 supposed Egyptian influence. There zodiac differs from the usual sequence. It starts at are Egyptian parallels for a divided zodiac, but, sig- the top with the bust of Aries, runs down to Virgo nificantly, none for a divided zodiac ring nor for a in the left half, then turns back to the top to Li- change of sequence of zodiacs. bra and runs down to Pisces in the right half (FIG. Indeed, another interpretation seems to be pos- 2). To explain this peculiarity the two halves have sible as the new photographs in Markoe 2003: figs. been interpreted as the civil year beginning with 197-199 allow a better interpretation. Not only the month Nisan on the left and the normal year is the sequence of the signs unusual, but also the iconography of various signs as busts (for a list of Nabataean busts at Petra cf. Wenning 2004). The zodiac is organized in pairs, the most prominent ones being the two on top. Aries and Libra indicate the spring and the autumn equinox, and East and West. In Babylonian mythology the ram is con- nected with the moon, the scales with the sun (cf. Papke 1993: 62, 200, 234), and their moon and sun are male deities. That might explain why Aries is here pictured as a male bust, wrongly interpreted by Glueck 1965: 415-416 as bust of Athena (op- posed by Starcky 1968: 231-232). At first glance it may seem far-fetched to bring in Babylonian models as an explanation, but one must consider that Hellenistic-Roman world was thoroughly acquainted with Babylonian astrono- my. That’s true for the Nabataeans too, as at least a few pieces of evidence demonstrate. Convincingly, the two masks in the frieze of the Tomb with the Armour (Br. 649) at Petra and the masks of some tombs at Hegra are interpreted as a representation of the beheaded Sumerian-Babylonian demon Humbaba (McKenzie, Reyes and Schmidt-Colinet 1998). I know of another head of Humbaba in a figural capital found at Petra, still unpublished. It must have been the intention to bring Aries and Libra together at the top of the zodiac, other- wise Pisces would have been next to Aries. This intervention into the sequence of the signs resulted in the unusual two halves of the zodiac. There is a second emphasis in the composition. The new arrangement of the signs of the zodiac brings Virgo and Pisces to the bottom of the ring as counterpart to the pair at top. While Aries and Libra as moon and sun indicate cosmic features, the sphere above 1. Completed relief of Nike with Zodiac. ‘Ammån, Jordan Archaeological Museum and Cincinnati Art Museum. the earth, Virgo and Pisces are related directly to Markoe 2003: fig. 199, Courtesy the Cincinnati Art Mu- the earth itself and its fertility. Virgo holds ears of seum. grain (cf. the Semitic name for Virgo “ear”, Arabic

-578- The Message of the Khirbat at-TannØr Reliefs

2. Zodiac with bust of the local female deity as Tyche. Cincinnati Art Muse- um. Markoe 2003: fig. 198, Courtesy the Cincinnati Art Museum. “sunbul”; kind information by F. Villeneuve) and struct such a frieze. A solution came only with the the fishes symbolize the living water. frieze of the Qaßr adh-Dharπ˙ with the alternation Therefore, I believe that the composition does of zodiac busts and standing Nike figures and gave not follow two calendars, but expounds local as- the idea for such a reconstruction of this frieze as pects of agricultural fertility of the land as a gift well, suggested first by Laurent Tholbecq. The re- by the local deities within a larger cosmological construction of the façade of the temple by McKen- conception. Some further arguments can confirm zie (McKenzie, Gibson and Reyes 2002: 56-65 fig. this interpretation. In the middle of the zodiac ring 13) is convincing (FIG. 3), although the position of there is the bust of a female deity, capite velato and the deities in the frieze is debatable. Three busts of with mural crown, a Tyche as tutelary deity. Behind planetary deities are placed in the middle intercol- her left shoulder projects a double sceptre crowned umnium of the frieze, two busts at the outer inter- with the cone of pine and the crescent moon. It is columnia. Each bust is framed by Nike figures, add- not a semeion as it is often called. Another crescent ing up to ten Nike figures. These unframed blocks moon is added in the field opposite above her right belonging to the frieze are discussed by McKenzie shoulder. The identity of the deity is explained (McKenzie, Gibson and Reyes 2002: 59-63; McK- below, as it is revealed thanks to the frieze of the enzie, Reyes and Gibson 2002: 460-461) and do temple — what matters now is that the double cres- not require more detail here. The best-preserved cent moon indicates the celestial reign of the de- blocks are those of Sol, Jupiter and Saturn. ity. Embedded into the zodiac with its stars and the Beside the busts of the planetary deities and the pronounced position of Aries and Libra, the monu- Nike figures there are four busts in framed panels ment reflects a cosmological programme. directly above the half-columns. These four panels allow further consideration of the composition. The Frieze The two middle panels were identified by Glueck as The frieze of the temple of Khirbat at-Tannør rep- Mercury (Glueck 1965: 228, 467-468 pls. 146a-b), resents a programme that closely corresponds to but one is representing Apollo with a small kithara this. Glueck 1945: 184 suggested a frieze of the behind his left shoulder (Starcky 1968: 233; McK- seven planetary deities, but was not able to recon- enzie, Gibson and Reyes 2002: 60), the other the

-579- Robert Wenning

3. Elevation of façade of inner temple enclosure, Period II (reconstruction J. McKenzie). Markoe 2003: fig. 180, Courtesy the Cincinnati Art Mu- seum. moon god (FIG. 4), as the rest of the crescent moon Hadad-Jupiter; 396, 411 pl. 53a, Tyche with lyre; above the shoulders proves (cf. Roche 1990: figs. McKenzie, Gibson and Reyes 2002: 59-60). Again, 4-5). This panel seems to be unfinished. Unlike to a clarification of the shown deities is necessary. On the moon goddess (Luna) among the planetary dei- the left a bust of Zeus is shown with a thunderbolt ties, this is a male moon god as in the composition behind his left shoulder (FIG. 5). He is to be com- of the zodiac. Apollo is a god of light representing pared with the male deity of the relief of the cult the sun, a well-known feature for this god. He is the figures where this deity is depicted as a mixture counterpart of the moon god. That means, sun and of the Syrian Hadad and the Greek Zeus (Glueck moon are again emphasized. The two outer panels depict the deities of the temple itself (Glueck 1965: 411 pls. 56, Zeus-

5. Panel from the frieze, Bust of the local male deity in a Zeus 4. Panel from the frieze. Bust of the moon god. ‘Ammån, type. ‘Ammån, Jordan Archaeological Museum. Photo by Jordan Archaeological Museum. Photo by R. Wenning, R. Wenning, Courtesy Department of Antiquties, ‘Am- Courtesy Department of Antiquties, Amman, Jordan. mån, Jordan.

-580- The Message of the Khirbat at-TannØr Reliefs

1965: pls. 41-42; McKenzie, Gibson and Reyes 74; McKenzie 2003: figs. 195, 193) and has been 2002: 74; Markoe 2003: fig. 40). The cult figure identified by Glueck as Atargatis, based on the Syr- accompanied by two bulls follows rather Syrian ian type of the cult figure seated on a throne accom- models, while the bust in the frieze is closer to panied by lions. But since the Syrian “Hadad” is Greek models. But there are connections between probably not Hadad, the Syrian “Atargatis” could the two. For instance, both wear unusual torques easily represent a local deity other than Atargatis. around their necks. The local deity behind these Nevertheless, Atargatis as an eastern appearance of Zeus types is probably Qos, to whom a dedication Hera remains a possibility. Lucian, De Dea Syria was discovered on site (Savignac 1937: 408-409 32 describes the Dea Syria as a very complex syn- no. 2; McKenzie 2003: fig. 196). Although this is cretistic construct (cf. Hörig 1979). If one prefers the only clear indication of a venerated deity at Kh- to call the female deity of Khirbat at-Tannør Atar- irbat at-Tannør beside the sculpture at all, and al- gatis, one has at least to modify the common inter- though it could be, that this dedication has nothing pretation. She is not the Fish- or Dolphin Goddess, to do with the deities of the temple at all, it makes and not necessary an Aphrodite type, but a matron, sense that the sanctuary at Khirbat at-Tannør could a tutelary deity, and a goddess of fate and fertility. have been devoted to the Edomite Qaus whose cult But for now one should refrain from labelling her continued among the Nabataeans and Idumaeans as at all. Instead of speculation about names the iden- Qos through the Hellenistic-Roman period. Hadad, tification of the female deity as a Hera type is of Qos, Zeus and even Dushara all more or less share great importance and must define the direction of the same iconographic type of weather and fertility interpretation for the moment. deities. The same couple of Zeus and Hera type deities The counterpart of Zeus in the frieze on the is depicted once again in two smaller panels of the outer right depicts a female deity capite velato and entrance either to the temple or to the Temenos holding a double sceptre (FIG. 6). She can be rec- (Glueck 1965: 467 pl. 153b, Hermes-Mercury, and ognized as Hera, the wife of Zeus. Like Zeus she 207 pl. 45a, Tyche) and underlines the given inter- bears a torques. Hera is much less common in the pretation. The Zeus this time is identical with the Levante, but when she is pictured she sometimes Jupiter of the planetary deity from the frieze (the holds ears of grain and points to the fertility of the same type is found in Petra in Markoe 2003: fig. 3; country as the female deity of this temple. The fe- Roche 2001: 355 no. 18), while the Hera is identi- male paredros of the Zeus type deity is depicted in cal with the Hera of the framed corner panel. Once a smaller relief which could reflect the cult figure again, the same local goddess in the Hera type is to again as a Syrian-Greek composite (Glueck 1965: be identified in the middle of the zodiac discussed pls. 44, 161b; McKenzie, Gibson and Reyes 2002: above. The panels of the corner pilasters of the frieze are decorated with a bust at the outer long sides as well, which is unusual. Here a Tyche figure with the horn of plenty (FIG. 7) is depicted (Glueck pls. 55, 53b, Dionysus; McKenzie, Gibson and Reyes 2002: 59-60). Such reliefs with Tyche figures are common in Nabataean art, often as reliefs above the framing half-columns. Here the number of planetary deities, Nikes, Zeus and Hera, and Apollo and the moon god left no space to show them up in their usual position. These figures do not represent the Tyche in the sense of a city goddess as they lack the particular attributes needed for such an identification and they are not temple deities, but rather symbolise the blessing, wealth, fertility, and 6. Panel from the frieze, Bust of the local female deity in protection guaranteed by the venerated temple dei- a Hera type. ‘Ammån, Jordan Archaeological Museum. ties. For convenience, I call them Tyche figures. It Photo by R. Wenning, Courtesy Department of Antiquties, is important to note that they are placed next to the ‘Ammån, Jordan. -581- Robert Wenning by Hera-Tyche as the local tutelary deity with astral elements.

The Tympanum Beside the frieze one also needs to mention the so- called Atargatis panel, a semicircular panel above the entrance of the temple (Glueck 1965: 143-144 pls. 31-33; McKenzie, Gibson and Reyes 2002: 63- 64; Markoe 2003: fig. 41), which depicts the bust of a female figure surrounded by floral motifs (FIG. 8). Following Glueck the figure usually is identified as the temple deity, which depicts Atargatis as a vegetation goddess, but I would like to sug- gest another interpretation. The body of the figure is covered with floral branches, which are directed 7. Panel from the frieze, but at lateral side, Bust of a Tyche figure. ‘Ammån, Jordan Archaeological Museum. Photo towards the bottom indicating that the figure origi- by R. Wenning, Courtesy Department of Antiquties, ‘Am- nates from the earth or even below. Even the neck mån, Jordan. and the face are covered with leaves. Such features do not fit Atargatis, but are characteristic of marine panels with Zeus and Hera, the temple deities. creatures in Greek art. There are two other Naba- Clearly, the arrangement of the blocks wit the taean busts of the same type known, one from Kh- temple deities send a message. The temple deities irbat Bråq (Parr 1960: 134-135 pl. 15,1), the other frame the frieze of the planetary deities and par- one from Petra (Lyttleton and Blagg 1990: 277 pl. ticipate in this way in the cosmic ideas. The local 9). If one conceder that Khirbat Bråq is important deities are related to the fertility of the herds and mainly for its spring, the ‘Ayn Bråq, which pro- of the land, while the planetary deities promise its vides water for Petra, this bust should represent the perpetuation. The same programme is expressed in personification of that spring. The Petra bust (FIG. the zodiac: the busts of Aries and Libra together 9a-b) could represent the ‘Ayn Møså or the Wådπ with the zodiac form the astral sphere, Virgo and Møså, but having no context for the bust this re- Pisces symbolise the fertility of the country, joined mains still hypothetical. At Khirbat at-Tannør one

8. Relief from the tympanon panel with the bust of the personification of the ‘Ayn La‘bån (?). ‘Ammån, Jordan Archaeological Museum. Markoe 2003: fig. 41, Courtesy the Cincinnati Art Museum.

-582- The Message of the Khirbat at-TannØr Reliefs

must have been strong relations, and probably pro- cessions took place between the two sites. Khirbat at-Tannør could have been a place of pilgrimage. The figure in the tympanum is not the temple deity and not a deity in the sense of the Zeus and Hera type deities, but rather comparable with the Tyche figures or the “Rankenfrau” in the tympanum of the al-Khazna and the Florentinus Tomb. The crowning element of the pediment, an eagle (Glueck 1965: pl. 34a), is taken from the Greek world as a symbol of Zeus, and in the Se- mitic world as a symbol of the god of heaven under the local name of Dushara, Ba‘al-Shamin or Qos. In the combination with the personification of the spring this feature connects the fertility of the earth with the local deity and the divine sphere. At least four times we find the same idea formulated among the Khirbat at-Tannør reliefs. The emphasis of astral and cosmic connotations at the sanctuaries of Khirbat at-Tannør and Khir- bat adh-Dharπ˙ is remarkable. One wonders if the programme reflects the new political situation of the Nabataeans after AD106. Although planetary deities do appear in earlier Nabataean art, even in connection with Nikes, there seem to be no earlier Nabataean representations of the zodiac. On the other hand, signs of zodiac are pictured on seals of the Roman administration in Moab at this time, found at Mampsis in the Negev. The end of the Nabataean kingdom by the Ro- man occupation and transformation into the Pro- vincia Arabia in AD106 must have shocked the Nabataeans. But rather than a break, one can ob- serve increased building activity now sponsored by various local authorities. This is to be understood as a revival of the particularistic structures of the 9a-b. Petra. Architectural panel with the bust of the personi- Nabataean tribal society, which Rabbel II had tried fication of a spring (?). Photo by R. Wenning, Courtesy to overcome in his re-novatio. At the two sites in Department of Antiquties, ‘Ammån, Jordan. northern Edom discussed here the Nabataean tradi- tion remained very strong, especially in the archi- would rather prefer to point to ‘Ayn La‘bån than to tecture, while the extremely rich sculptural decora- the Wådπ al-Óaså. I would like to identify the bust‘ tion was a rather new feature. The people running as the personification of the ‘AynL a‘bån. There is a the sanctuaries still worshipped their traditional Nabataean inscription from the site, dated to 8/7BC, deities. While these gods formerly protected the which mentions a votive by Netir’el, the master of royal dynasty, now they had risen to a transcen- the spring of La‘bån (Savignac 1937: 405-408 no. dental astral and cosmological sphere. In such tur- 1). This obviously important spring is situated at the bulent times, the builders, artists and worshippers foot of Khirbat adh-Dharπ˙, and the Wådπ La‘bån apparently chose to pin their hopes on astral deities continues towards Khirbat at-Tannør. Between the rather than seeking the favour and protection of the isolated sanctuary on Jabal at-Tannør and the settle- gods of their new Roman overlords, who started to ment at Khirbat adh-Dharπ˙ 7km to the south there gain popularity elsewhere, especially in the cities.

-583- Robert Wenning Acknowledgments McKenzie, J.S., Reyes, A.T. and Gibson, S. 2002. Khir- Gratefully I acknowledge permissions and bat at-Tannur in the ASOR Nelson Glueck Archive support by the Department of Antiquities of Jor- and the Reconstruction of the Temple. Annual of the dan and its director general, Prof. Dr. Fawwaz al- Department of Antiquities of Jordan 46: 451-476. Khraysheh, and Dr. Glenn Markoe, The Cincinnati McKenzie, J.S., Reyes, A.T. and Schmidt-Colinet, A. Art Museum, and helpful comments by J. McKen- 1998. Faces in the rock at Petra and Medain Saleh. zie, F. Villeneuve, M.-J. Roche, M. B. Reeves, J. Palestine Exploration Quarterly 130: 35-50. P. Oleson, A. E. Knauf, and A. Lichtenberger. A. Papke, W. 1993. Die geheime Botschaft des Gilgame- Kropp kindly checked the English, remaining mis- sch. Bergisch Gladbach: Lübbe. takes are mine. Parr, P.J. 1960. Nabataean Sculpture from Khirbet Brak. Annual of the Department of Antiquities of Jordan Bibliography 4-5: 134-136. Glueck, N. 1945. The Other Side of the Jordan. New Roche, M.-J. 1990. Bustes fragmentaires trouvés à Pé- Haven: ASOR. tra. Syria 67: 377-395. ––– 1952. The Zodiac of Khirbet et-Tannur. Bulletin ––– 2001. Figurines, Sculpture, and Reliefs. Pp. 350- of the American Schools of Oriental Research 126: 358 in Z. T. Fiema et al. (eds.), The Petra Church. 5-10. American Center of Oriental Research Publications ––– 1965. Deities and Dolphins. The story of the Naba- 3. Amman: ACOR. taeans. New York: Farrar, Straus and Giroux. Rosenthal-Heginbottom, R. 2001. A Nabatean Zodiac. Gundel, H.G. 1992. Zodiakos. Tierkreisbilder im Alter- Pp. 51-54 in O. Rimon and R. Shchori (eds.), Measu- tum. Kulturgeschichte der antiken Welt 54. Mainz: ring and Weighing in Ancient Times. Haifa: Hecht von Zabern. Museum. Hörig, M. 1979. Dea Syria. Studien zur religiösen Tra- Savignac, R. 1937. Le dieu nabatéen de La‘aban et son dition der Fruchtbarkeitsgöttin in Vorderasien. Alter temple. Revue Biblique 46: 401-416. Orient und Altes Testament 208. Kevelaer: Butzon Starcky, J. 1968: Le temple nabatéen de Khirbet Tannur. & Bercker and Neukirchen-Vluyn: Neukirchener. Revue Biblique 75: 206-235. Lyttleton, M.B. and Blagg, T.F.C. 1990. Sculpture from Villeneuve, F. 2002. Une louve romaine à sabots chez the Temenos of Qasr el-Bint at Petra. Aram 2: 267- les Nabatéens?, Pp. 93-94 in L’archéologie à l’École 286. normale supérieure. Ouvrage dédie à Christian Pey- Markoe, G. 2003. Petra Rediscovered. Lost City of the re. UMR 8546 CNRS-ENS. Archéologies d’Orient Nabataeans. New York: Abrams. et d’Occident. Paris. McKenzie, J.S. 2001. Keys from Egypt and the East: Villeneuve, F. and Al-Muheisen, Z. 2000. Nouvelles Observations on Nabataean Culture in the Light of recherches à Khirbet edh-Dharih (Jordanie du sud, Recent Discoveries. Bulletin of the American Schools 1996-1999). Comptes rendus de l’Académie des in- of Oriental Research 324: 97-112. scriptions: 1525-1563. ––– 2003. Carvings in the Desert: The Sculpture of Pet- ––– 2003. Dharih and Tannur, Sanctuaries of Central ra and Khirbet et-Tannur. Pp. 165-191 in G. Markoe Nabataea. Pp. 83-100 in G. Markoe (ed.), Petra Re- (ed.), Petra Rediscovered. Lost City of the Naba- discovered. Lost City of the Nabataeans. New York: taeans. New York: Abrams. Abrams. McKenzie, J.S., Gibson, S. and Reyes, A.T. 2002. Re- Wenning, R. 2004. Nabatäische Büstenreliefs aus Petra construction of the Nabataean temple complex at – zwei Neufunde. Zeitschrift des Deutschen Palästi- Khirbet et-Tannur. Palestine Exploration Quarterly na-Vereins 120: 157-181. 134: 44-83.

-584-